为了将决策者对各目标属性的模糊评判信息转换为目标偏好信息,首先将模糊语义转换为三角模糊数,利用模糊数的广义加法、近似乘法和标量乘法进行计算,从而将决策者对目标属性的离散意见转换为对各目标的综合意见;然后采用多指标模糊排序法确定决策者权重,通过定义一种模糊综合排序指标来确定各目标偏好权重,依据目标权重构建判断多目标Pareto解的适应度函数,并采用粒子群算法对多目标问题进行求解;最后通过一个算例来说明该算法的实用性和有效性.%In order to transfer fuzzy evaluation information of objective' s attributes into objective preference information, this paper transferred fuzzy semantics into triangular fuzzy number, and used generalized additive function, an approximate multiplication and a scalar multiplication between fuzzy numbers to compute.Then transferred discrete levels of objective' s attributes to integrated levels, and used fuzzy sorting multi-indexes to define weights of method makers.After that, defined a fuzzy integrated indexes to determine each objective sorting weight, defined a fitness function of multi-objective Pareto solutions based on these objective weights,and used particle swarm algorithm to solve the multi-objective problems.Finally, used a case to illustrate the algorithm's feasibility.
展开▼
